1. Field of the Invention
The present invention is generally directed to loudspeaker systems and more particularly to loudspeaker systems which use sound chambers which progressively propagate entering annular mid frequency sound waves concentrically about high frequency sound waves to an output wherein the mid frequency sound waves are substantially parallel on opposite sides of the high frequency sound waves.
2. Brief Description of the Related Art
Most loudspeaker systems for commercial or professional applications require more than one transducer. There are two common reasons for this that stem from the limits of transducer technology: limited bandwidth; and/or limited sound power output of individual transducers.
The limited bandwidth of transducers, when compared with the wide bandwidth of the human ear dictates the need for multi-way loudspeaker systems. The wavelengths of sound audible to us range from nearly sixty feet to less than three quarters of an inch in length. No single transducer can reproduce this range of frequencies with acceptable levels of both distortion and efficiency.
The limited sound power capacity of a single multi-way loudspeaker unit when compared to the sound power and distribution required for large venues, dictates the need for multi-unit loudspeaker groups or arrays. This is the case in nearly all commercial use or professional loudspeaker systems. For the purposes of this discussion, multiple units of multi-way loudspeakers will be considered.
Clarity, referred to also as intelligibility and speech intelligibility, is affected by the degree to which the loudspeaker reconstructs the temporal and spectral response of the reproduced wavefront. Interference in the perception of that wavefront can be caused by environmental reflections of sound waves bearing the same spectral information which arrive near in time to the beginning of the wavefront.
Coherence of a wavefront refers to the degree to which the loudspeaker reconstructs the temporal response of the reproduced wavefront.
Uniformity of distribution refers to the similarity in the temporal and spectral nature of the reproduced sound when considered spatially.
Correction of the sound spectrum through equalization is easily achieved with signal processing equipment. Correction of the temporal aspects of sound referred to as impulse response equalization is considerably more complex. Correction of the spatial distribution of sound energy, after the sound has exited the loudspeaker system is not possible.
To fully understand all aspects concerning clarity in large loudspeaker systems, it is necessary to consider issues beyond those limited to the temporal and spectral performance of individual transducers and their related enclosures or waveguides. Wavefront coherence and uniformity must be considered concerning several aspects of the multi-way structure and the multi-unit array. In the multi-way loudspeaker the additional issues are twofold; the reconstruction of complex waveforms from two or more transducers not physically occupying the same location that reproduce different parts of the spectrum; and the temporal interference that occurs in the region of spectral overlap between transducers. In the multi-unit array a further consideration is added: the temporal interference between multiple transducers working together to reproduce the same part of the spectrum.
Complete and uniform energy summation occurs when two or more simple cone loudspeakers produce sound waves of the same frequency which propagate into the same space, where the wavelength propagated is approximately equal to or greater than the spacing of the loudspeakers. In cases such as this the devices are said to be mutually coupled; multiple devices work nearly as a single device.
Complex patterns of summation result in reduced spatial uniformity and lost efficiency when two or more transducers produce sound waves of the same frequency which propagate into the same space, where the wavelength propagated is smaller than the spacing of the transducers. These patterns are not easily integrated in systems and most often, the result is reduced coherence of the wavefront and therefore reduced sound quality.
It is evident that a useful approach to the problem of summation is to physically limit or eliminate the negative interaction between adjacent transducers through the design of wavefront modifying or directivity controlling mechanical geometry through which the sound waves are propagated. The mechanical control of such interactions are therefore of great interest in the development of better loudspeaker arrays.
From the ideal loudspeaker system, sound would appear to the listener as though it came from a point source floating in space. This goal is approachable in a single multi-way loudspeaker, but impossible in a large sound system. Nevertheless, audio engineers have sought over the years to come as close to the goal as possible through a number of interesting innovations.
In small systems, it can be said generally that for best coherency, the physical spacing between transducers of differing frequency ranges should be kept as small as possible. Whereas in large systems, more attention should be paid to the physical relationship between transducers operating in the same frequency range due to the overall size of the array.
The evolution of the co-axial loudspeaker has resulted in improved coherency in two-way systems. A typical variation is a two-way device consisting of a high frequency compression driver mounted on the back plate of a woofer magnet, so configured to allow the sound from the high frequency driver to pass through the woofer and emerge at the center of the cone of the woofer. The passageway through the low frequency magnet combined with the woofer cone, or other small horn device, serve to guide the high frequency energy. The addition of time compensation in the signal path to correct for the physical displacement of the two sound sources produces something very close to the ideal. In this described configuration a direct radiator is combined with a horn loaded driver.
However, the directivity cannot be controlled to the extent that might be desired at all frequencies in such a loudspeaker. Furthermore, a substantial part of the benefit of point source approximation is lost when multiple co-axial speakers are configured in an array spaced on the centers of the woofer. The larger size of the woofer may result in the space between high frequency drivers increasing beyond the dimension allowed by the smaller high frequency drivers, thus aggravating the interference problem between the high frequency components. It is evident that the co-axial driver can improve coherence in a small system, but where large multiples are deployed, no significant gain is likely to occur.
The recently introduced co-entrant horn disclosed in U.S. Pat. No. 5,526,456 to Heinz is a two way, mid frequency and high frequency horn loaded variation on the co-axial loudspeaker. In this variation, the high frequency compression driver is mounted on the back plate of a mid frequency compression driver magnet, so configured to allow the sound from the high frequency driver to pass through the mid frequency device and emerge through the center of the diaphragm of the mid frequency driver. The energy from the mid frequency diaphragm enters the throat of the horn through an annular slot adjacent to the high frequency opening. With suitable time compensation to align the acoustic output of the two devices in the time domain, the result is similar to the co-axial loudspeaker, but with the added advantages of increased mid frequency efficiency and control of mid frequency directivity through the horn loading of that band of energy. However, the discontinuity in the high frequency throat caused by the mid frequency entrance to the throat of the waveguide is quite close to the high frequency driver diaphragm. If the discontinuity is within one quarter wavelength of a given frequency, energy reflected back to the diaphragm will arrive at the half wave interval fully out of phase and cause disruptions in response.
The improvement in the relationship between the two elements within the device, is offset by increased spacing between the high frequency drivers in an array caused by the size of the mid frequency horn. In large arrays therefore, no improvement in high frequency coherence or uniformity of distribution is likely to occur.
Coherency in loudspeaker arrays is a far more complex problem than that of coherency in the single multi-way loudspeaker. Firstly because of the potential size and number of elements to be found in arrays and secondly because of the more difficult acoustic environment and listener configuration in which arrays are typically applied.
Large numbers of transducers are required in large and small auditoria, compounding the problems of spatial distribution and coherence. Where the system design specifies such loudspeakers to be widely distributed throughout the environment, the state of the art with respect to loudspeakers seems sufficient.
Wide distribution throughout the listening space is generally not acceptable where a large public sound system is oriented to music or speech performance. The acoustical focus of the audience, is in most cases, the stage. It is then a primary requirement that an array of multiple speaker enclosures will be placed in close proximity with one another in front of and facing the audience in order to complement that focus. Generally there are at least two arrays of loudspeakers flanking the stage. It is equally inevitable that the interactions between loudspeakers within each array will play a significant role in the outcome.
The consideration of wavelength is preeminent in the science of sound: all sound phenomena are at least in some aspect wavelength dependent. Design considerations with respect to loudspeaker interaction in large arrays are in fact dominated by consideration of wavelength.
First, the wavelength of any frequency under consideration in the array will determine in which frequency range the individual transducers are coupled with one another and in what range they are interfering. Secondly, the directivity of any device is wavelength dependant; the directivity will determine the degree of angular overlap of adjacent wavefronts and therefore the degree of potential acoustical interference.
Wavelength variation of three orders of magnitude over the audio spectrum assures us that no one transducer can possess the same radiation characteristics over the whole audio spectrum. In fact, even when the spectrum is divided into three separate frequency ranges, most transducers operating even within these reduced bandwidths demonstrate a continuous change in the radiation pattern of their acoustic energy with changing frequency.
While a phenomenon can be useful in one frequency range, it may be detrimental in another. One effect, destructive interference, is generally just that. However, this phenomenon can also be used to limit unwanted energy beyond the edge of an area of desired coverage, such as with a di-pole radiator.
Another effect, mutual coupling, while generally regarded as a positive with respect to efficiency and wavefront coherence, can also be a hindrance when beam width narrows excessively. Coupling between drivers, combined with electrically induced phase shift is also responsible for the undesirable effect of beam tilting through the crossover region between two drivers. Mutual coupling occurs when drivers are placed within approximately one wavelength of one another. See Olson, Elements of Acoustical Engineering, 1944 Van Nostrand and Co.
In a line array (Olson et al) in its simplest form, a row of closely spaced direct radiators, is dependant on mutual coupling of one driver to the next. Historically, line arrays have consisted of multiple small direct radiating transducers arranged in a vertical row. Typically the drivers are chosen to be sufficiently small to allow mutual coupling to the highest frequency of concern. For example four inch diameter drivers permit coupling to above 3 Khz, which is sufficient to allow good speech intelligibility. This approach yields a system with a controlled vertical coverage and correspondingly wide horizontal coverage.
Another variation on the line array is a vertical column of high frequency compression drivers mounted on horns with narrow vertical beam width. However, the mutual coupling is limited to a small portion of the lower range of the high frequency transducer.
The ribbon tweeter can be considered a line array of nearly infinite elements, with all the attendant benefits. However, limits in sensitivity and power handling capacity have not permitted the ribbon tweeter to replace the preeminent position of the high frequency compression driver in systems for large spaces.
Spatial distribution of energy within the listening environment has increasingly become the focus of efforts by practitioners of the audio arts. The result of this effort is a number of novel innovations.
Very old established principles which define the line source of Olson et al. are now being combined with significant new trends including new geometry for the purpose of modifying high frequency wavefronts. See for example U.S. Pat. Nos. 5,163,167 to Heil and 5,900,593 to Adamson.
In the interests of improved coherence, spatial distribution and frequency response, a number of high power, high fidelity line array variations have recently been introduced. These multi-way systems all approach the different frequency bands with different technology. While most of these new concepts rely on prior art in the direct radiator portion of the array, several new concepts have emerged in the effort to create line arrays to the highest discernable frequency.
The prior art patents to Heil and Adamson reveal high frequency acoustic sound chambers (that are sometimes referred to as waveguides) capable of wavefront transformation to the highest audio frequencies, for use with compression drivers and waveguides. The output of such devices provide an essentially continuous ribbon of coherent high frequency sound. When placed end to end, even in large arrays, high frequency coherency is maintained. This high frequency solution is seen in curved horizontal and vertical arrays in Adamson and flat vertical arrays in Heil.
Other high frequency sections of new line arrays consist of a previously described simple vertical row of conventional high frequency horn and driver units.
In the mid frequency range significant unresolved problems are apparent. Two general categories of solution are now in use: horn loaded and direct radiator systems. The benefits and limitations of these solutions must be considered with respect to vertical and horizontal arrays.
When direct radiators are used in a mid frequency vertical array, it is not regarded as a suitable solution to place a single mid frequency line array beside a high frequency array. The lack of horizontal symmetry will result in undesirable variations in frequency response across a horizontal section of the array. A more likely solution is to place two vertical line arrays spaced equidistant from a central high frequency line array.
However, due to upper frequency requirements of the mid frequency direct radiators, a maximum size limitation is imposed. This size limitation is incompatible with the demand for substantial acoustic power in the mid band. In such applications, the direct radiating mid frequency devices cannot match the acoustic output of the more efficient high frequency combination of waveguide and compression driver.
Furthermore, the horizontal spacing between the two vertical line arrays of mid frequency devices introduces a special set of limits due to the behavior of the two sound sources. When the two line arrays are spaced at the half wavelength of a given frequency, the energy from one line array arrives at the other 180 degrees out of phase and a cancellation of energy occurs. At higher frequencies the wavefront is divided into a number of narrow lobes due to variable summation between the two sources. While some control of directivity is achieved the gain is offset by losses due to the cancellations, which further reduce the efficiency of the direct radiators.
Much higher efficiencies can be achieved with horn loaded mid frequency, but the typical horn loaded horizontal or vertical arrays results in significant increases in driver to driver spacing. In such systems the mid section behaves as a coupled line array only in the lower half of the spectrum handled by the transducer. Above that frequency the array performs somewhat like a row of point source radiators with all the associated patterns of interference.
When the mid frequency is horn loaded in two columns placed symmetrically about the high frequency array, off axis problems arise due to the differing acoustic centers of the midrange and high frequency arrays. These problems arise due to the physical size of such devices.
In the case of three-way systems where a low frequency section is employed, there are few problems with conventional horizontal and vertical line arrays since these long wavelengths permit mutual coupling with conventional 12xe2x80x3, 15xe2x80x3 and 18xe2x80x3 woofers in the appropriate frequency ranges. Acoustic efficiencies and wavefront shape present few problems.
The present invention is comprised of a plurality of loudspeaker enclosures arranged in a horizontal or vertical array, where each enclosure must contain at least one high frequency compression driver and at least one inner sound chamber similar to that disclosed in U.S. Pat. No. 5,163,167 to Heil or as disclosed in U.S. Pat. No. 5,900,593 to Adamson or other high frequency throat piece as required to connect a high frequency driver to a waveguide, and at least one mid frequency driver and at least one outer mid frequency sound chamber so shaped to substantially enclose the inner high frequency sound chamber within the mid frequency sound chamber, whereby the inner surface of the outer sound chamber and the outer surface of the inner sound chamber form an acoustic passageway whose input orifice is annular and whose output orifices approximates two parallel slots of approximately uniform width which may be curved or flat. The enclosure may contain an extension of the high frequency sound chamber and the mid frequency sound chamber to further direct the sound waves after the exit of the sound waves from the high frequency and mid frequency sound chambers.
Where the loudspeaker enclosures are arranged in a vertical array the vertical cross section of the enclosure may be trapezoidal or rectangular and where the loudspeaker enclosures are arranged in a horizontal array the horizontal cross section of the enclosure may be trapezoidal or rectangular.
In the present invention there are no differences in principle or geometry between a horizontal array and a vertical array. The horizontal array is a simple 90 degree transformation of the vertical array and vice versa. Depending on the desired application, various embodiments may be constructed and oriented in any desired angle to suit the desired application.
In the typical embodiment the high frequency driver is fixed to the back plate of the magnet assembly of the mid frequency driver and is so placed to be concentric with and axially aligned to the mid frequency driver and the high frequency sound chamber is aligned axially and affixed concentrically to the front side of the mid frequency magnetic assembly which is so constructed to allow high frequency sound to pass through the magnetic structure of the mid frequency driver and to enter into the entrance of the high frequency sound chamber.
The mid frequency sound chamber is fixed to the front side of the mid frequency driver and is so placed to be concentric with and axially aligned to the mid frequency driver and is so shaped to form at least one passageway which is defined by the outer surfaces of the outer walls of the high frequency sound chamber and the inner surfaces of the inner walls of the mid frequency sound chamber with the at least one passageway extending from the annular input orifice to the rectangular output orifice of the mid frequency sound chamber.
The at least one passageway may be divided into at least two passageways which extend the full length of the high frequency sound chamber extending from the annular input orifice to the rectangular output orifice so configured to divide the annular input orifice into at least two arc segments and to shape the output orifices as two equal and parallel rectangular slots, defined by the outer surface of the high frequency sound chamber and the inner surface of the mid frequency sound chamber. A further aspect of the present invention is that the outer surface of the high frequency sound chamber and the inner surface of the mid frequency sound chamber provide a smooth and continuous transition in the cross sectional shape of the passageways to permit a gradual transformation of the shape of the mid frequency wavefront from an arc segment at the entrance to rectangular at the exit.
In the preferred embodiment, the outer surface of the inner high frequency sound chamber is modified to assist in the smooth transition from the annular input orifice to the rectangular output orifice. To facilitate this, a wedge shaped body of material is added to the sides of the high frequency sound chamber so shaped that the thin edge of the wedge divides the annular input orifice into two arc segments. The wedge shaped body of material expands in width as the distance from the input orifice increases thus changing the shape of the passageway according to the width of the wedge.
Furthermore in some embodiments the wedge shaped body is flattened and tapered in thickness and so shaped to conform to the inner surface of the mid frequency sound chamber to provide mating surfaces whereby the outer surface of the high frequency sound chamber is fixed to the inner surface of the mid frequency sound chamber.
In the preferred embodiment the outer surface of the inner high frequency sound chamber is extended at the output orifice to provide an additional high frequency acoustic load and to further guide the high frequency sound wave in a beam width of the desired angle. The outer surface of the inner sound chamber is further modified to provide a smooth passageway for the mid frequency sound wave propagated in the outer sound chamber as it passes out from the output orifice of the outer sound chamber.
A further aspect of the present embodiment is that the dimension of the outermost width of the dual rectangular output orifices of the mid frequency sound chamber is limited to less than one wavelength of the highest frequency that is expected to be propagated solely by the mid frequency sound chamber. The mid frequency sound chamber is therefore capable of propagating a wavefront into the cabinet waveguide to which it is connected to the highest frequency of concern without undesired narrowing of the beam width. Because of the close proximity of the two mid frequency exits, the mid frequency energy appears acoustically at the center of the waveguide. Because the exit of the high frequency sound chamber is located in the center of the two mid frequency sound chamber exits and thus at the center of the waveguide, both the mid and high frequency sound appear to originate acoustically from the same location. This geometry can be extended in a line, vertically or horizontally, with as many devices as required. An array of such sound chambers can be considered therefore, to be co-linear.
In the present embodiment the co-linear exit of the mid frequency and high frequency sound chambers is preferably joined to the entrance of the waveguide constructed according to the teachings of Adamson, U.S. Pat. No. 5,900,593 or according to the practice of Heil, U.S. Pat. No. 5,163,167.
In some embodiments, the enclosure may contain one or more low frequency loudspeakers, which may be configured to radiate sound in any manner which is deemed acceptable to provide the required low frequency sound power to complement the mid frequency and high frequency drivers.
Another distinct aspect of the preferred embodiment is that acoustical interference is created at the exits of the mid frequency sound chamber and the high frequency sound chamber due to discontinuities in reflected impedance and acoustic cancellations. These negative effects occur where the sound waves merge at the entrance to the waveguide, and are limited to a controlled bandwidth.
The interference is caused because the mid frequency wavefront encounters a discontinuity in acoustical resistance due to the space occupied by the high frequency sound chamber exit. Likewise, the high frequency wavefront encounters a discontinuity in acoustical resistance due to the space which is occupied by the exit of the mid frequency sound chamber. Both these discontinuities cause acoustical reflections and cancellations which result in degraded frequency response. These discontinuities are encountered by either the high frequency or mid frequency wavefront when propagated in the absence of the other wavefront and the frequency of the interference is dependant on the dimensions of the sound chamber exits.
In the preferred embodiment, the discontinuities of the passageways of both frequency bands are so sized that the interference occurs in a frequency range in which both high frequency and mid frequency drivers are capable of full acoustic output. The solution to the interference is found in time alignment of the mid frequency and high frequency wavefronts and the overlap in the frequency domain of the two frequency bands of sound. The result of this is that a transducer operating at a frequency where destructive interference will occur when the driver operates in the absence of the other frequency band does not encounter any interference when both drivers are operated simultaneously. This is so because the exits of both the mid frequency and high frequency sound chambers and thus the entire entrance of the waveguide is acoustically energized in the frequency range of concern.
An object of the present invention is to provide a method to create at least two wavefronts of at least two frequency ranges within a loudspeaker enclosure which will merge within the loudspeaker enclosure to form a single wavefront with virtual zero interference that includes all the acoustical energy of both wavefronts and both frequency ranges.
It is a further object of the present invention to provide a method to allow at least two wavefronts of a common frequency range and at least two wavefronts of a another common frequency range to produce a common wavefront within the same loudspeaker enclosure.
It is a further object of the present invention to provide a method to create one or more wavefronts within one or more loudspeaker enclosures that will merge with the wavefront(s) of the same frequency range in an adjacent similar loudspeaker enclosure with virtually zero interference.
It is a further object of the present invention to provide the optimal transformation of the shape of a sound wave between the exit of a mid range compression driver and the entrance of the associated waveguide by means of particular sound chambers.
It is a further object of the present invention to provide a method to eliminate interference between two wavefronts of different frequency ranges at the point of summation at the exit of particular sound chambers and the entrance of the associated waveguides by the application of particular geometric shapes, time delay and particular filtering of the sound signal in the electronic domain.